Making effective comparison charts

If you’re in marketing, then sooner or later you’ll end up making a comparison chart of some sort or another. They’re on the backs of packages, in literature, and on web sites. It’s a popular form of selling. But the kind you put in front of customers aren’t the type of charts I’m writing about here.

Comparison charts for internal use

Instead I’m writing about are the comparison charts we use internally – for product planning, competitive analysis, sales force education, and other reasons. These often get short shrift because they tend to get whipped together in a hurry and their very nature makes them hard to create and hard to maintain. It’s the kind of job that makes a lot of folks groan.

In this article I hope to give you some tips that will make them more effective, easier to create and easier to maintain. It doesn’t have to be all drudgery, and a well-made comparison can be a very serious asset.

Just so we’re clear about what we’re talking about, a comparison chart for the purposes here is a grid, a matrix. Across the top are products, companies or whatever is being compared. Down the side are the attributes that form the comparison. In the cells that make up the body, the individual attribute values for each entity being compared.

The attributes are the key

The goal is to use every attribute that has a reason to be on the chart, but not one single extra attribute. Adding attributes that don’t matter to the equation just makes life complicated, and invites complexity that isn’t needed.

Don’t just ask sales for the attributes

For sure you have to ask sales – they can tell you what the customers value. But they are not to the only source. Almost everyone in the company can (and should be asked to) suggest attributes, but there is one source that will take a bit of work: lost orders. Very often lost orders are quickly attributed to price, but after a little digging, other reasons come out as well. Put these on the list.

A good filter will narrow the list

Now comes the hard part, and that is filtering the list of attributes. This is also where the chart’s ultimate purpose needs to be decided, because a chart made purely for sales will likely contain fewer attributes than a chart made for marketing/product management/engineering use.

Here’s a list of criteria for keeping an attribute on the list:

  • The attribute has won or lost a sale by itself.
  • The attribute has been used in competitor’s ads/marketing materials, and has been mentioned by customers as a decision factor.
  • The attribute has an established, defined, measurable, demonstrable, value to the customer, i.e. it saves x% of expenses.
  • The attribute is highly correlated to either sales volume or price, or some other important metric, even if you can’t say why.
  • The attribute is unique, or highly variable. Use this criteria only for experimental charts, or charts being used for exploratory purposes.

If you are making the chart only for sales education, take only the top three

Ok, you’ve probably narrowed the list down a bit – it’s crucial to do that first, because there is nothing more depressing than keypunching values that don’t seem to have any value.

Define each Attribute, and why it is on the list

The next step is to write down why each attribute is on the list – which criteria, and why it applied. Then, write a definition for each attribute. This may seem silly – I mean, how confusing is “Weight.” Wait, was that tare weight? Empty weight? Curb weight? Gross weight? You get the idea.

Phrasing the attributes

Ultimately, when the attribute values are being filled in for each entity, you want to be entering one of three things: Either a yes or no, one of a defined list of values, or a number. Some may disagree (sorry Matt!) but this is critical for making a clear chart, making a search, sort and calculation-friendly chart, and making sure that the attributes are fully distilled and understood. It may mean that an attribute has to be broken down into several attributes, but that’s ok. Things will be clear, and that’s what we’re after. This rule may be bent for final versions of education-only charts, but for analysis it should be followed.

First try to make the chart as-is, and then go back and decide whether you could add a new entity or product without creating new value types – new numbers are ok. The real goal here is to completely avoid subjective language, inconsistent terms, and other entries that will become distractions when people use the chart.

Write definitions for any values that aren’t clearly self-explanatory. This applies especially to pseudo-specific terms like large, medium and small.

Now, make the chart. But how? The obvious choice is a spreadsheet, but it’s not always the best choice. Here are some things to consider:

  • If the chart is a working document, will be maintained only by one person, and there is a some calculation required, use a spreadsheet.
  • If the chart is about education, and you’ll need to make it look really professional, make a working copy in a spreadsheet but do the final chart in something like Adobe’s Illustrator.
  • If the chart is a working document, you have a lot of attributes, and plan to do a lot of analysis, use a database. Some kinds of analysis you can do: Correlating attributes with each other. Finding combinations of attributes that do and do not exist. making histograms of attribute values. All of these can be really useful, but are pretty ugly to do frequently in a database.
  • If the chart is one of many that have to be made available to a large group of people, consider using competitive intelligence/marketing software like Strategy Software’s Strategy! and IntoAction – some of the best CI software out there.

So you’ve picked the tool you’re going to use, and now you’ve got to fill it in. Where do you get the info? That’s a subject for another day 😉 For values where you don’t have a good source, fill in your best guess. Mark it somehow so that you know it’s a guess. For values you don’t have any source, don’t guess. Leave it blank. Any honest chart is likely to have a blank or two, and just make it clear to people that empty space means “we don’t know.” A side benefit is that it encourages people to write values in, which often lead to sources.

After you’ve got the chart filled in, it’s time to go back to the folks who fed you information, suggestions for attributes, and who will be using the chart. They’ll probably complain that one thing or another isn’t on the chart – pull out the criteria and explain them. They’ll probably ask about a few values, which your definitions will make clear. If there are any serious flaws or errors, fix them. Check your guesses – people naturally jump at the chance to correct things they think are wrong, but if no one does, ask. If no one has an answer you have a choice to make – either leave it in (with a footnote – don’t mislead people) or take it out.

Now get it to the right people, in the right way 

Do they need an electronic version? Paper? Does it need to be laminated? Maybe even wallet sized? The chart that you’ve spent a lot of effort making won’t be used if it can’t be used easily, so make it available easily.

Bask in the glory – you’ve done an unpopular job methodically and well, and people will appreciate it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s